Non-invasive methods to image tumor metabolism have outstanding potential and could in principle allow the clinician to predict tumor genetics, stage, likelihood of therapeutic response and other crucial parameters that have direct bearing on patient care. However, the forms of metabolic imaging currently in use - particularly FDG-PET and low-field magnetic resonance spectroscopy - focus on general hallmarks of malignancy and do not provide any specific molecular information about the tumor. Here we propose to develop a method to predict the genotype of gliomas based on in vivo imaging of the "oncometabolite" 2-hydroxyglutarate (2HG). This metabolite accumulates only in gliomas that bear specific mutations in the genes encoding isocitrate dehydrogenase-1 and -2 (IDH), and the presence of these mutations in gliomas is strongly associated with low- grade lesions and a favorable prognosis. We will develop a proton magnetic resonance spectroscopic imaging (1H-MRSI) technique at 3T for detection of 2HG in the human brain and to determine noninvasively in vivo if the IDH mutations occur in patients with gliomas. Specifically, the research team will investigate spectral editing techniques, including scalar difference editing and multiple-quantum filtering. MRSI sequences will be first designed using computer simulations and will be validated in phantoms and healthy brain in vivo. Next, a clinical study in brain tumor patients will be carried out in patients with low grade gliomas or secondary glioblastomas enrolled in our clinical protocol. Evaluation of the IDH1 status by immunohistochemical (IHC) analysis of resected tumor will be performed and the patients will undergo 1H-MRSI to assess 2-HG status. The results will be correlated with IDH1 mutational status. We will study 30 patients of which we anticipate ~20 will have the mutation. An additional 15 patients with glioblastoma will be studied as negative control subjects since IDH1/2 mutations have not been found in this patient group. This work will lead to the first method to detect a tumor-specific metabolite in vivo. Furthermore, because our approach is easily integrated into existing 3T MR clinical systems and does not require radioactive probes, we anticipate few barriers to its translation into general practice. While this research will facilitate the utility of the proposed technology in the study of gliomas, it should be stressed that the technique will be widely applicable to other diseases in which alterations of 2HG levels are implicated.